Recognition of Ancient Tamil Brahmi Symbols and Notations Using ELM-ISMO Algorithm

Digitalizing ancient images, symbols and notations from stone inscriptions which reveals the cultural history is a way to preserve and emphasize our country’s heritage. This paper discusses, symbols from ancient Tamil Brahmi scripts are recognized using Extreme Learning Machine (ELM) with Improved Spider Monkey Optimization (ISMO) algorithm. Initially, the input images are preprocessed using Histogram Equalization (HE) method followed by segmentation using KMeans clustering algorithm. Then, the segmented characters are recognized and classified into equivalent Tamil characters using ELM-ISMO algorithm. The performance results of ELM-ISMO algorithm is compared with ELM, CNN and ANN algorithms. The performance metrics sensitivity, specificity, accuracy and F1-Score are measured. Performance results have shown that the proposed ELM-ISMO outperforms the existing algorithms in terms of accuracy, sensitivity, specificity and F1-score metrics.